S. C.Blanchard and N. D. Chasteen
1362
Electron Paramagnetic Resonance Spectrum of a Sea Shell. Mytilus edulis Stuart C. Bianchard and N. Dennis Chasteen. Department of Chemistry, University of New Hampshire, Durham, New Hampshire 03824
(Received January 26, 1976)
Publication costs assisted by the University of New Hampshire
The EPR spectra of the periostracum, the calcitic prismatic region, and the aragonitic nacre of the shell of the three layer bivalue Mytilus edulis were investigated. The organic periostracum exhibits a g = 4.3 signal due to high-spin Fe3+ in a low symmetry coordination environment and a g = 2.0044 f 0.0002 signal presumably due to a semiquinone or quinone radical. However, the Fe3+ signal can account for only a small fraction of the approximately 0.1% total iron by weight, in this tissue. The prismatic region exhibits a weak Fe3+ signal a t g = 4.3 and strong Mn(I1) lines arising from manganese substituted for calcium in the calcite lattice of the shell. Examination of the Mn(I1) spectrum of an oriented section of the prismatic region reveals a high degree of structural organization with the principal axis of the zero field tensor paralleling the iridescent cross sectional striations (long axis of the prisms) which correspond to the c axis of the shell. Organic material remaining after decalcification of the prismatic region with HCl displays a radical signal, g = 2.0022 f 0.0002. The nacre exhibits the g = 4.3 Fe3+ signal and a doubled g = 2.0044 f 0.0002 radical signal. Oriented sections of the nacre have a very weak but anisotropic Mn(I1) signal of unknown origin. The concentration of Mn(I1) in the calcite of the prismatic region, as determined by EPR, correlates positively with the position of the animal above low tide. These results suggest that EPR might be useful in investigations of calcified tissue, particularly the exoskeletons of marine animals.
Introduction
structure and mucopolysaccharide which serves as a cement. I t is generally believed that the organic matrix provides sites of nucleation for CaC09 crystals and dictates the type of crystalline polymorph that form^.^^^ The blue mussel is an example of a typical three-layer bivalve mollusk (Figure 1).loThe outer layer of the shell or periostracum is largely composed of quinone tanned protein. The deep blue crystalline layer just beneath the periostracum is the prismatic region which contains unique anvil-like prisms of calcite.1° Finally, the inner pearl-like nacreous layer consists of aragonite crystals arranged in neatly packed horizontal rows. In the crystalline portions of the shell, the individual crystallites are surrounded by matrix protein.1° Biomineralization activity is centered in the extrapallial fluid and the mantle (Figure l).4We have examined in some detail, the spectra of the individual components of the shell of Mytilus
The electron paramagnetic resonance (EPR) spectra of transition metal ions doped into synthetic minerals have been extensively investigated over the years;' these studies have been valuable in establishing crystal field effects on the electronic structure of metal ions in various ligand environments. However, the spectra of paramagnetic species naturally occurring in mineralized tissue have received little attention. In particular, the spectra due to radicals and transition metal ions found in the exoskeletons of marine organisms have not been examined. We have become interested in the possibility of using EPR spectroscopy to obtain information concerning the oxidation states and coordination environments of trace paramagnetic transition metals in calcified t i ~ s u e .I~t is, ~not known whether these impurities occur in unusual environments or occupy specific lattice sites which possess symmetry relative to the external morphology or growth characteristics Experimental Section of marine shells. If they do, conceivably these paramagnetic EPR spectra were measured on a Varian E-9 spectrometer metal ions could serve as probes of the microstructure of sea operating at X-band and 100-kHz magnetic field modulation. shells and other calcified tissue. Oriented specimens were mounted on a quartz rod attached Metal ion speciation is an important aspect of modern anto a goniometer. A rectangular dual cavity was employed with alytical chemistry. The application of EPR to examine minVarian strong pitch, g = 2.0028, as a reference. eralized tissue could be of use in investigations of biominerSamples of fresh Mytilus were collected on May 15,1975, alization processes and of biogeochemical cycling of trace elfrom five levels above low tide on the beach near Odiorne State ements as well as in other fields. We report here a model study Park, N.H., and labeled NRO1-NR05. They were frozen until of the EPR spectra of the shell of the edible blue mussel, Mytilus edulis. needed. Single shells for Mn and Fe atomic absorption analysis were Sea shells are composed of 97-99% CaC03 (calcite, aragocleaned in 2.5% NaOH to remove the periostracum as prenite, or vaterite) with lesser amounts of MgC03, (Al,Fe)203, SiOa, Ca3P208, CaS04, protein, and mucop~lysaccharides.~-~viously described.3 The periostracum was prepared for atomic absorption analysis by dissolving about 0.1 g in several milliIn addition to these major and minor constituents, trace liters of concentrated nitric acid with gentle heating. The soamounts of Sn, Mo, Mn, Cd, Ti, B, Pb, Au, Ag, Ni, Co, Bi, Cu, lution was diluted to 25.0 ml with l N HC1. A 5.0-ml aliquot Sr, Rb, and As have also been found in varying amount^.^-^ was removed and diluted to 25 ml with 1 N HC1 for analysis. Investigations of the structure of mollusk shells often reveal A Varian Techtron Model AA-3 spectrometer modified with a highly ordered arrangement of organic and mineral comAA-5 electronics was used for the atomic absorption analysis. p o n e n t ~Tiny . ~ crystals are separated from one another by an Wavelengths (and slit widths) employed were 279.68 nm (100 organic matrix consisting of a protein with a keratin type The Journal of Physical Chemistry, Voi. 80, No. 12, 1976
1363
EPR Spectrum of a Sea Shell
PERIOSTRACUM
1
PR'SMATl$
M y t i l u s edulis
I
I
MANTLE -PRISMS
~ N V I L -L I K E
snbPEo
--NACRE
coMmsE0
CF H O R I Z O N T A L ROWS
I
S H E L L OF M Y T I L U S EDULIS (NOT D R A W N TO S C A L E )
Figure 1. Schematic diagram of the shell structure, extrapallial fluid, and mantle for the edible blue mussel, Mytihs edulis.
g 34.3 1000 G
2000G
3000 G
Figure 2. Liquid nitrogen temperature EPR spectrum of the periostracum of Mytilus.
and 248.33 nm (50 p ) for Mn and Fe, respectively. Additional details are given e l ~ e w h e r e . ~ The prismatic and nacreous layers were separated by scraping or breaking off pieces from the shell. A particularly large specimen was used to obtain samples for the orientation studies. Sufficient quantities of periostracum for EPR and metal analyses were removed from about 12 shells by scraping and were dried for 10 min at 50 "C. The mineral form of the calcium carbonate in several powdered samples was determined by x-ray diffraction on either a General Electric Model BR1 or a North American Phillips unit both using Cu KCYradiation.
p)
Results and Discussion A. Periostracum. The EPR spectrum of the organic periostracum a t 77 K is shown in Figure 2. Two principal resonances occurring at g = 4.3 and g = 2.0 are observed. The sharp resonance near g = 2.0 is probably due to an organic radical since the room-temperature power saturation behavior is similar to that of other radicals such as melanin found in hair (Figure 3). The g value for the radical is 2.0044 f 0.0002 with a peak-to-peak line width u = 6.9 f 0.2. For an older sample, g = 2.0047 with u = 6.7 G was observed. Resonance lines with g values of 2.0046 were also observed for the periostracum of the Atlantic Ribbed Mussel Vossella demissa (u = 9.2 G) and the Razor Clam, Solen Viridis (u = 10.1 G). Quinone and semiquinone radicals which exhibit EPR signals are frequently associated with various pigments in living systemsll and typically have g values in the range 2.0040 to 2.0050.12 T h e values for the periostracum of various species are consistent with the belief that this material is composed in part of quinone tanned protein^.'^ The resonance line near g = 4.3 (Figure 2) with u = 70 G and the associated line at g = 9.5 are characteristic of high spin Fe3+ in a completely rhombic e n i ~ r o n m e n t . *These ~ J ~ signals are frequently observed with non-heme iron proteins.14-'6 At room temperature, the spin-lattice relaxation time, 7'1, is considerably shorter and the iron signal is only 1/22 as intense as that at 77 K (Figure 2). The iron signal can be adequately described by the S = 5/2 spin Hamiltonian 1
+
1+
+
- - S ( S 1) E (S,' - S y 2 ) gpA S (1) [ 3 in which D and E are the axial and rhombic components, respectively, of the zero field. In the situation giving rise to the g = 4.3 resonance E / D = 1/3.14 The total iron content of the periostracum was determined 7-f = D S,
Periostracum Melanin
0 0
5
10
(POWERP hw)1/2
1
Figure 3. Room-temperature power saturation curves for the g = 2.0044 f 0.0002 peak of the periostracum and the free radical mel-
anin. by atomic absorption to be 12 500 ppm. The high concentration of iron in the periostracum is probably indicative of an important role of iron in this tissue which has not been noted p r e v i ~ u s l y . ~A~semiquantitative J~ determination of the spin concentration from the EPR spectrum indicates that the signal in Figure 2 accounts for less than 10%of the total iron in the sample, the remaining iron being EPR silent a t 77 K. The lack of an EPR signal might be attributed to an unusual structure consisting of exchanged coupled Fe3+ ions or possibly Fez+ ions which are difficult to observe by EPR.' The intensity of the g = 4.3 resonance changes very little when the periostracum is soaked in six changes of 1%(w/v) o-phenanthroline, pH 4.5, over several days until no more of the reddish iron o-phenathroline complex is observed. Apparently, the iron responsible for the EPR signal binds tenaciously. However, the total iron content of the periostracum is significantly reduced to only 707 ppm, or 5.7%of the original value of 12 500 ppm. This indicates that the EPR silent iron is bound relatively weakly. Additional studies are clearly indicated. B. T h e Prismatic Region. A t 77 K, a g = 4.3 iron signal is also observed with the prismatic region although it is much less intense than observed for the periostracum. The iron content for this portion of the shell is typically only 20 to 50 ppm. In this case, the iron is probably coordinated to the matrix protein of the shell. The Journal of Physical Chemistry, Vol. 80, No. 12, 1976
S. C. Blanchard and N. D. Chasteen
1364
A singlet EPR signal was observed a t g = 2.0022 f 0.0002 with u = 10.5 G for the organic material obtained from a HC1 decalcified section of the prismatic layer. The origin of this signal is not known; however, it is likely due to a free radical. The g = 2 region of the spectrum of a powdered sample of the prismatic region is shown in Figure 4. The spectrum is typical of Mn2+ in carbonate rninerals.I8 The prismatic region contains typically 10 ppm Mn. The six main lines arise from the coupling of the electron magnetic moment, S = 5/2, with the manganese nuclear moment, Z = 5/2, for the M s = -1/2 +1/2 transitions. The other transitions such as M s = +1/2 +* -3/2 are very anisotropic and often difficult to observe with powdered samples. The weaker pairs of lines between the main lines are the so-called forbidden lines in which both electron and nuclear spin states change, i.e., AMs = 1and AMI = 1 (strong field notation).Ig As we shall see, the EPR data are consistent with Mn(I1) substituted for Ca(I1) in the calcite lattice of the prismatic region. Figure 5 shows the metal site of trigonal symmetry in the rhombohedral unit cell of calcite. The threefold axis is coincident with the crystallographic c axis. The axially symmetric spin Hamiltonian without quartic terms in the electron spin is given by
PISMA MA TIC
MANGANESE (11) M y t i l u s edulis
I
-
r
-
+
sin H cos O/g2)2(1/2Ho)(4S(S 1)
- 24Ms(Ms - l ) - 91 - (DgL2 6'/g2)(1/8HO)12S(S + l ) - GMs(Ms - 1) - 31 - d(l40Ms" - 210Ms2 + [190 - GOS(S + 1)lMs + 30S(S + 1) - 60}[35(gllCOS 0/g)4 - 3O(gll COS O/g)2 + 31 + KMI - [ R 2 ( A 2+ K2)/K2](1/4H~)[Z(Z + 1) - MI^ + M1(2Ms2 - l)]- (D2AM,)(1/8H02)~[6Ms(Ms - 1) - 2S(S + 1) + 31 sin4 0 + [3Ms(Ms - 1) - 2S(S + 1) 1
- [S(S
-
+ l)l2/[Ms(Ms- l)]]sin2 20)
The Journal of Physical Chemistry, Vol. 80, No. 12, 1976
Figure 4. Room-temperature EPR spectrum of the g = 2 region for the powdered prismatic region of Myfilus edulis. At liquid nitrogen temperature a g = 4.3 resonance is observed.
(2)
where A and R are the parallel and perpendicular hyperfine constants, respectively. The general equation for the resonance fields of allowed transitions in a strong magnetic field using perturbation theory to second order is given in the classic paper by Bleany.20 Hurd et a1.21have expanded this general equation by including a d' term required in the interpretation of spectra exhibiting large axial crystalline-field effects, and also include fourthorder terms. However, they omit the important third-order term later given by Bleany and Rubins.22The third-order term is given as the energy of the spin state. From the energy, we have evaluated the third-order contribution to the resonance M s - 1 transition. When simplified, the field for the M s result is in agreement with the third-order term given by Tsay -1/2 transition. There has been et a1.23for the M s = 1/2 some question as to the correctness of various terms in the equation for the resonance fields.24 The general equation, which we believe to be correct, including third-order terms is given below. The following conventions apply. (1)A, R, and D are defined to have negative valueslg and are expressed in gauss. Thus the M I = -5/2 line for the Mn(I1) spectrum is the low field line. (2) D has been substituted for D/2 in the terms taken from Hurd et a1.21to be consistent with the nomenclature of later papers.
+ (Dg,gll
3600 G
3300 G
G
1
+ ASJ, + B(S,Z, + SyZy)+ D 1 S z 2- 3 S(S + I ) ]
-
I 3000
(3)
1
0
I
Carbonate
Figure 5. The Mn(ll) site in a calcite lattice. 0 denotes oxygen atoms from different carbonate anions.
where
A' =
R' = Bg,P K' = l/g[(A'gll cos
+ (R'g,
sin 0)2]1/2
K = K'/gP g = [(gl,cos 6')2 u
+ (g,
(4) sin 6')2]1/2
= gmark PHmark l h
Ho = h ulgP Here h is Plank's constant and 6' is the angle between the symmetry axis and the applied magnetic field. Equation 3 is applicable to oriented samples. For powders, 1/2 D can be obtained from the doubling of the M s = -1/2 hyperfine lines due to second and higher order terms in the perturbation equations. These lines are isotropic to first order. The complicated general equations describing these effects22 can be reduced to
-
for the Mn(II), S = 5/2, system under consideration. Here a is the hyperfine splitting measured from the powder spectrum. a and D are expressed in gauss and are taken to be negative.lg The value of 6 H is usually taken from the M I = 5/2 (high-field line) where the splittingof the doublet is maximum.The spectrum in Figure 4 is characterized byg = 2.0006 & 0.0002, a = -93.8 f 0.2 G, and D = -86.3 & 0.2 G2sfrom eq 5 . These values are in reasonable agreement with those published for Mn(I1) in calcite." We were particularly interested in establishing whether the
1365
EPR Spectrum of a Sea Shell
M A N G A N E S E (11)
1I
1
I
I
1
0"
SINGLE O R I E N T E D S E C T I O N P R I S M A T I C MyiWuS eduhus
I
PARALLEL
I f
0
30
90 120 ANGLE LB.Degrees)
60
0- mark
150
180
Figure 7. Variation in line position for the MI = +5/2 Mn(ll) line of the single oriented section of the prismatic region
NACRE 90' P E R P E N D I C U L A R 3100
G
3400G
3700G
Figure 6. The Mn(ll) spectrum at three angles for section of the prismatic region of Mytilus.
a single oriented
Mn(I1) EPR spectrum could be used to establish the degree of order of the individual crystallites of the shell. The long dimensions of the prisms in the prismatic layer are oriented with varying angles relative to the shell surface depending on the section of the shell (see Figure 1).loBecause of the difficulty in doing electron microscopy and EPR spectroscopy on the same shell section, we attempted to estimate the orientation of the prisms from the shell color. Careful examination of the prismatic region of a shell from a large animal revealed iridescent striations which varied in angle with the shell surface in the same way as the long axis of the prisms in Figure 1. A section of the shell in which the striations made an angle of about 45O with the shell surface was chosen to search for any anisotropy in the EPR spectrum. The shell section was mounted on a quartz rod such that the striations made an angle 8 with the magnetic field vector. The spectrum as a function of angle is shown in Figure 6. The six +1/2 transitions are evident. The variamain M s = -1/2 tion in the intensities of the forbidden lines with angle are similar to those found for Mn(I1) in c a 1 ~ i t e .Note l ~ that the forbidden lines gain intensity a t the expense of the allowed lines. A t certain orientations, other transitions which are very -312 can be seen (arrows in anisotropic such as M s = +1/2 Figure 6). Unfortunately, these transitions could not be observed over a sufficiently large range of angles to permit their close examination. The positions of the M I = 5/2 line varies by about 16 G over the range of angles due to second- and third-order terms in D. The observed and computer calculated line positions using eq 3 are shown in Figure 7 . In the calculation, we assume that the direction of the shell striations and the principal axis of the magnetic tensor (c crystallographic axis) are coincident. The calculation is for the parameters gii = 1.9998 f 0.0002, g, = 2.0004 f 0.0002,A = -94.1 f 0.2G,R = -94.0f0.2G,and D = -80.0 f 0.2 G which were obtained from the spectrum to give the best overall fit to the observed line positions. Travis'O found through a lengthy electron microscopic and diffraction study that the c axis of the calcite crystallites paralled the long dimensions of the prisms in which they were located. She estimated that the prismatic region was composed of nearly 100% highly ordered crystalline calcite. The
-
-
Myt i Ius edulis
Iu 1 g = 2.00463 I
32806
3330G
3380
Figure 8. EPR spectrum showing the splitting of the g = 2.0 peak for the nacre.
good agreement between the observed and theoretical line positions (Figure 7) and the sharpness of the spectrum (Figure 6) are in accord with this. Not all shells exhibit such a high degree of order as Mytilus. The common barnacle, Balanus balanoides, of which practically nothing is known about its shell structure, has been examined by EPR in a similar fashion.26The EPR spectrum indicates that only about 75% of the calcite microcrystallites possess any degree of order; the rest of the shell consists of polycrystalline substance in which much of the Mn(I1) is in sites of low symmetry. In this case, the crystallographic c axis of the ordered calcite crystallites is normal to the shell surface. C. The Nacre. The only clearly observed resonance for powdered samples of the nacreous layer was a relatively sharp line (u = 10 G) a t g = 2.0044 f 0.0002 shown in Figure 8. The doubling of the line suggests that two radical species are present. This resonance may be closely related to the g = 2.0044 radical observed for the periostracum. A t liquid nitrogen temperature, a weak Fe(II1) signal becomes visible a t g = 4.3. No Mn(1I) spectrum was observed for the powdered nacreous layer a t room temperature or liquid nitrogen temperature although the sample contained about 8 ppm Mn by atomic absorption spectroscopy. However, when an oriented section of the nacre is placed in the cavity, a weak anisotropic Mn(I1) signal is observed (Figure 9). (The sloping baseline is due to Fe203, g = 2.1, trapped between the growth layers of the shell.) The EPR parameters g = 2.0026, a = -95 G, and The Journal of Physical Chemistry, Vol. 80, No. 72, 1976
S. C. Blanchard and N. D. Chasteen
1366
TABLE I: Metal Content of Prismatic Region" Sample
3:OO G
3400 G
3700 G
Flgure 9. The EPR spectrum for a single oriented section of t h e nacre of Mytilus. The normal to the shell surface is perpendicular to the magnetic field vector.
D
= -80 G are comparable to those for Mn(I1) in various
carbonate minerals.ls We were not able to locate a perfect symmetry axis such that rotation about this axis produced an invariant spectrum, although the principal axis of the zerofield tensor appears to be roughly perpendicular to the underside of the shell. The metal site could have rhombic symmetry. The nacre was shown by x-ray diffraction to be aragonite with no detectable calcite; thus the Mn spectrum is not due to Mn(I1) in calcite (or calcite impurities in aragonite). However, the signal is not likely due to Mn(I1) substituted for Ca(I1) in aragonite either, since the signal is quite different from that which we have observed with aragonitic shells.27Its origin remains obscure. When a powdered sample of the nacre is heated for 1 h at 600 "C above the transition temperature (520 "C) from aragonite to calcite, strong Mn(I1) signals similar to those for the prismatic region were observed (Figure 4). X-ray diffraction confirmed the transition to calcite. It is not clear why most of the Mn(I1) exhibits no E P R spectrum in the nacre of the unheated shell. Experiments a t liquid helium temperature may prove useful. D. Correlation of t h e E P R Spectrum with Tidal Position. The trace metal ion composition of marine shells is sometimes found to correlate with their environment.28Table I shows the results for the metal analysis for the prismatic region of M y tilus for five animals collected at different positions above low tide. The concentration of Mn(I1) substituted for Ca(I1) in calcite (Table I) was determined by comparison with a standard sample and using the formula
where S,is the peak-to-peak first-derivative signal height of the unknown, G , is the instrument gain setting, W t , is the weight of the unknown in the sample tubes, and slope is the slope of the plot of the signal height divided by the instrument gain vs. sample weight for the known sample. The known sample was a barnacle (RN002) with U k = 8.30 f 0.02 G for the M I = -5/2 low-field line.3 For the prismatic region of Mytilus, nu = 7.68 f 0.02 (Figure 4). Equation 6 assumes the same line shape function for both samples, which is reasonable. Precautions required in this type of analysis are detailed elsewhere." There appears to be no relationship between the total M n by AA and the position of the animal in the intertidal zone (Table I). However, there is a strong positive correlation (coeff The Journal of Physical Chemistry, Vol. 80, No. 12. 1976
Heightb AAppmFe
AAC ppm Mn
EPRd ppm Mn(I1)
NROl 0.0 19.4 8.6 4.7 0.5 42.2 10.6 4.1 NR02 NR03 1.0 59.3 12.0 8.2 NR04 1.5 47.3 11.5 7.8 NR05 2.0 27.4 10.6 11.1 a Error in values estimated to be &3%. Approximate height above low tide in meters. Total manganese by atomic absorption spectroscopy (method of standard additions). Based on the EPR signal vs. weight curve for reference sample RN002 with line width correction. Mn(I1) substituted for Ca(I1) in calcite. See ref 3.
-5
I
10-
:-
a
v
z 8-
0
0
z
0 4 00 10 20 APPROX I M A T E H E I G H T ABOVE LOW T I D E ( M E T E R S )
Figure 10. Correlation of the concentration of Mn(ll) in calcite of the prismatic region of Mytilus against the height of the animal above low tide (0.0). The linear regression line is shown. 0.916), albeit on a limited number of samples, between the Mn(ZZ)substituted for Ca(ZZ) in calcite and the tidal position (see Table I and Figure 10). Interestingly, a similar correlation has been observed for the total M n in barnacle ~ h e l l s . ~ * ~ ~ I t would appear that E P R might be useful in investigating environmental influences on the accumulation of manganese in different structural components of marine shells. Conclusion The results presented here suggest that EPR spectroscopy can be profitably used to examine certain trace elements and radicals in calcified tissue and to obtain information concerning structural organization. E P R might be employed in organic fractionation studies to help identify which components of the periostracum or organic matrix are responsible for the observed radical signals. Biomineralized tissue can provide unusual metal ion environments in which to study crystal field effects on transition metal ions. For example, the Ca(I1) ion in aragonite is nine c o ~ r d i n a t e Marine .~~ animals appear to be one of the few sources of this thermodynamically unstable form of CaC03. Surprisingly, a study of Mn(I1) doped in aragonite has never been reported. We find that aragonitic shells exhibit a very unusual Mn(I1) spectrum in which the forbidden lines are more intense than the allowed lines at X-band frequenciesz7
Acknowledgment. The authors wish to thank the National Science Foundation, Grant No. NSF M P S 75-03474,and the University of New Hampshire Central University fund for support of this research.
Lattice Vibratlons of Quinhydrone References and Notes
1367 (15) W. T. Oosterhius, Struct. Bonding (Berlin), 20, 59 (1974). 1161 and W. E. Blumbero. "Metallooroteins as Studied bv Electron ~ J., Peisach Paramagnetic Resonance" in';Electron~Spin Resonance of Mhai Complexes", T. F. Yen, Ed., Plenum Press, New York, N.Y., 1969, p 72. (17) V. R. Meenakshi, P. E. Hare, N. Watabe, and K. M. Wilbur, Comp. Biochern. Physiol., 29, 6 1 1 (1969). (18) T. R. Wiideman, Chern. Geo., 5, 167 (1969). (19) P. A. Narayana, J. Chem. Phys., 55, 4263 (1971). (20) B. Bleany, Phil. Mag., 42, 441 (1951). (21) F. K. Hurd, M. Sachs, and W. D. Hershberger, Phys. Rev., 93, 373 (1954). (22) B. Bleany and R. S. Rubins, Proc. Phys. SOC.(London), 77, 103 (1961); corrigendum, 78, 778 (1961). (23) F. Tsay, S. L. Manatt, and S.I. Chan, Chem. Phys. Lett., 17, 223 (1972). (24) H.W. de Wijn and R. F. van Balderen, J. Chem. Phys., 46, 1381 (1967). (25) The stated errors reflect only the uncertainty in the magnetlc field measurements due to the line widths. They do not take into account any systematic error due to the possible noncoincidence of the principal resonance fields and the experimental line positions due to "powder effects". (26) S. C. Bianchard and N. D. Chasteen, manuscript in preparation. (27) L. K. White, A. Szabo, and N. D. Chasteen, work in progress. (28) See, for example. 0. H.Pilkey and R . C. Harriss. Limnol. Oceanogr., 11, 381 (1960); C. M. Gordon, R. A. Carr, andR. E. Larson, ibid., 15,461 (1970). (29) J. P. R . de Villiers, Am. Mineral., 56, 758 (1971). ~
For review see H. A. Kuska and M.T. Rogers in "Radical Ions", K. T. Kaiser and L. Levan, Ed., Wlley, New York, N.Y., 1968, pp 579-745; and B. A. Goodman and J. D. Raynor, Adv. lnorg. Chem. Radioctmm., 13, 135 (1970). E. A. Burgess, N. D. Chasteen, and H. E. Gaudette, Environ. Geol., 1, 171 (1975- 1976). S. C. Bianchard and N. D. Chasteen, Anal. Chim. Acta, 82, 113 (1976). K. M. Wilbur in Physiology of Mollusca", Voi. 2, K. M. Wilbur and C. M. Yong, Ed., Academic Press, New York, N.Y., 1964, p 243. K. M. Wilbur and K. Simkiss, Compre. Biochem., 26A, 229 (1968). E. T. Degens, D. W. Spencer, and R. H. Parker, Comp. Biophys. Biochem. Physiol., 20, 553 (1967). R. R. Brooks and M. G. Rumsby, Limnol. Oceanogr., I O , 521 (1965). P. Tasch, "Paleobiology -. of the Invertebrates", Wiley, New York, N.Y., 1973, pp 312, 881. S. Weiner and L. Hood, Science, 190, 987 (1975). D. F. Travis, J. Ultfastruct. Res., 23, 183 (1968). A. E. Needham, "The Significance of Zoochromes" in "Zoophysiology and Ecology", Vol. 3, Springer Veriag, New York, N.Y., 1974. M. Adams, M. S. Blois, Jr., and R. H. Sands, J. Chem. phys., 28, 774 (1958). R. E. Hiiiman, Science, 134, 1754 (1961). H.H.Wickman, M. P. Klein, and D. A. Shirley, J. Chem. Phys., 42, 2113 (1965).
~~~~
~
~
Lattice Vibrations of Quinhydrone and the Intermolecular Potential in the Crystal Kunio Fukushima' and Masataka Sakurada Department of Chemistry, Faculty of Science, Shizuoka University, 836, Oya, Shizuoka, Japan (Received December 23, 1975)
Infrared spectra of quinhydrone, quinhydrone-d2, and phenoquinone were measured in the region of 400030 cm-'. On the basis of the observed vibrational frequencies, the intermolecular potential in the quinhydrone crystal was investigated by carrying out calculation of the optically active lattice vibrations.
The crystal structure of quinhydrone has been studied by Matsuda e t al. 'and also by SakuraL2 According to their studies, each hydroquinone molecule in the crystal is connected to two quinone molecules by hydrogen bonds, and the perpendicular projection of the hydroquinone molecule on the neighboring quinone molecule seems to indicate that the interaction of the two kinds of molecules by charge transfer force is not so strong as other charge transfer complexes, for example, the hexamethylbenzene-chloranil complex. Lattice vibrations of the crystal are expected to reflect sensitively the strength of interactions due to the charge transfer force. Spectroscopic investigations of quinhydrone in the lower frequency region have not yet been made, although those in the higher frequency region have been carried o u t . 3 ~There~ fore, in the present investigation vibrational spectra of quinhydrone in the lower frequeiicy region were observed and the interaction of the hydroquinone molecule with quinone molecules in the crystal was studied by carrying out calculation of the optically active vibrations on the basis of the experimental results. Experimental Section Quinhydrone crystal was prepared by slow evaporation of a n acetone solution containing a n equimolecular mixture of hydroquinone (prepared by recrystallization of hydroquinone (grade G.R., Wako Pure Chemicals Co. Ltd.)) and p-benzo-
quinone (grade G.R., Wako Pure Chemicals Co. Ltd.) at room temperature. The crystal melted a t 167-168 "C. Phenoquinone crystal was prepared in a similar way from a petroleum solution containing phenol (grade G.R., Wako Pure Chemicals Co. Ltd.) and p-benzoquinone in molar ratio of 2 to 1. The crystal melted a t 70-71 OC. Quinhydrone-d2 was prepared by dropping 1 cm3 of a cooled heavy water solution containing 0.1875 g of sodium periodate slowly into 1.5 cm3 of a cooled heavy water solution containing 0.38 g of hydroquinone-d2 and then drying the generated precipitate. Infrared spectra of crystal samples in a Nujol mull and of phenoquinone in solution were measured. A Hitachi EPI-G3 infrared spectrophotometer was used for the measurements in the region of 4000-400 cm-', and a Hitachi FIS-3 far-infrared spectrometer for the region of 400-30 cm-l. Spectra of solutions were measured using KBr window cells of 0.1 mm thickness and polyethylene cells of 0.5 m m thickness. The results of measurements are shown in Tables I and 11. Calculation of Optically Active Lattice Vibrations Quinhydrone crystal formed by slow evaporation of an acetone solution containing hydroquinone and p-benzoquinone in a molar ratio of 1:1is monoclinic.'s2 The crystal structure has been analyzed as belonging to the space group P21/c.lv2 Two chemical units of C&&C&02 belong to each lattice point of the unit cell of the crystal. Constellation of the two The Journal of Physical Chemistry, Vol. 80, No. 72, 1976
